FORENSIC APPLICATIONS OF LA-ICP-MS: ELEMENTAL PROFILING AND EVALUATION OF HOMOGENEITY IN SODA- LIME CONTAINER GLASS Karen J. Harrington
Inductively Coupled Plasma Mass Spectrometry (ICP-MS) Previous method of RI predominantly used Elemental analysis research began three decades ago Drawbacks Allows classification of product type and use Shown to be a discriminating technique for forensic analysis Dangerous digestions Sample size requirements Destructive nature
Laser ablation (LA)-ICP-MS A viable method for forensic glass comparisons Creates aerosol of super-fine glass fragments Advantages Decreased sample prep. time Minimal sample consumption Maintains good sensitivity for most elements Less risk of contamination
Homogeneity Are elements consistent within a glass item (micro-homogeneity)? Natural heterogeneity (result of manufacturing) Contamination, natural variation in raw material How do elements vary between glass items? Variation beyond natural fluctuation?
Previous Research on Homogeneity LA-ICP-MS does reveal small natural variations in glass Understanding homogeneity within a unit of glass is a factor when: Comparing samples from same categorical sources Comparing samples from different manufacturers
Homogeneity Glass items typically maintain some homogeneity overall Glass of the same type can often be distinguished on an item-item basis Glass bottles tend to be more variable and may not be distinguishable from each other within one manufacturer
Project Goals How consistent are elemental profiles in glass bottles? What are appropriate match-criteria to balance false positives and false negatives? How variable are glass bottles within a manufacturer?
Background Stage I Verify valid set of match-criteria Examine homogeneity of a single bottle Determine what criteria allow for treatment of bottle as a whole unit
Background Stage II Determine normal/natural manufacturer variability Hourly runs Daily runs Simultaneous/daily for multiple furnaces
Method Sample Selection Gallo Glass Company, Modesto, CA Typical soda-lime, with cullet (25% recycled) Stage 1 10 bottles from two furnaces, variable time intervals, colors, types (3 fragments each from neck, sidewall, base regions of each bottle) Stage 2 Hourly, daily, daily simultaneous furnace
Method Analyte Selection Analytes were selected from an acid digestion of selected bottles from the population. Ratios were determined based on %RSD and mass-to-charge values.
Method Comparison Criteria T-tests and ANOVA were found to be unacceptable for the purposes of this project. Typical case-work match-criteria of ±2 SD was used
Method Instrumentation Perkin-Elmer ELAN DRC II ICP-MS (a) New Wave Research Nd: YAG LA (1 = 213nm) unit (b) a, b Dodds, A.J., Land, D.P., Pollock, E.M. Determination of Elemental Homogeneity in Automotive Windshields by LA- ICP-MS. The CACNews. 4 th Quarter 2005. 17 20.
Results Stage I 10 comparisons of multiple regions showed indistinguishable elemental profiles for 13 ratios with ± 2SD. When ± 3SD was applied to Sr/Zr and Fe/Cr (the two consistently problematic ratios), all 10 bottles showed indistinguishable elemental profiles for 15 ratios.
Results Stage I Comparison Areas Inside Neck vs. Sidewall Inside Sidewall vs. Base Inside Base vs. Neck Outside Neck vs. Sidewall Outside Sidewall vs. Base Outside Base vs. Neck Neck vs. Sidewall, all Sidewall vs. Base, all Base vs. Neck, all Inside vs. Outside Distinguishable Elemental Profiles Indistinguishable elemental profiles
Results Stage I Comparison Areas Inside Neck vs. Sidewall Inside Sidewall vs. Base Inside Base vs. Neck Outside Neck vs. Sidewall Outside Sidewall 4000.0 vs. Base Concentration Outside Base vs. Neck Neck vs. Sidewall, all Sidewall vs. Base, all Base vs. Neck, all Inside vs. Outside 5000.0 3000.0 2000.0 1000.0 0.0 #1 #11 Distinguishable Elemental Profiles Inside Neck vs. Sidewall, Ca/Mn #23 #29 #34 #47 Sample # Indistinguishable elemental profiles #58 #59 #70 #148 Mean (in) Mean (is)
Results Stage I Comparison Areas Inside Neck vs. Sidewall Inside Sidewall vs. Base Inside Base vs. Neck Outside Sidewall vs. Base Outside Base vs. Neck Neck vs. Sidewall, all Sidewall vs. Base, all Base vs. Neck, all Inside vs. Outside Concentration Outside Neck vs. Sidewall Distinguishable Elemental Profiles 6000.0 5000.0 4000.0 3000.0 2000.0 1000.0 0.0 #1 Indistinguishable elemental profiles Neck vs. Sidewall, Ca/Mn #11 #23 #29 #34 #47 #58 Sample # #59 #70 #148 Mean (N) Mean (S)
Results Stage I Comparison Areas Inside Neck vs. Sidewall Inside Sidewall vs. Base Inside Base vs. Neck Outside Neck vs. Sidewall Outside Sidewall vs. 1000.0 Base Outside Base vs. Neck Neck vs. Sidewall, all Sidewall vs. Base, all Base vs. Neck, all Inside vs. Outside Concentration 5000.0 4000.0 3000.0 2000.0 0.0 #1 Distinguishable Elemental Profiles Inside vs. Outside, Ca/Mn Indistinguishable elemental profiles #11 #23 #29 #34 #47 #58 #59 #70 #148 Sample # Mean (i) Mean (o)
Results Stage 2 (hourly) Comparison Areas Group 1 furnace 3 Group 2 furnace 2 Group 1 vs. Group 2 Distinguishable Elemental Profiles Indistinguishable elemental profiles * * * Group 1 (one bottle distinguishable for 1 ratio) and Group 2 (two bottles distinguishable for 3 ratios) show some distinguishable elemental profiles from other bottles.
Elemental Ratios Sr/Zr Na/Al Comparison Areas Ca/Mn Zn/Rb Fe/Cr Group 1 furnace 3 Ba/Pb Group 2 furnace 2 Ni/V Ti/Li Group 1 vs. Group 2 Results Stage 2 (hourly) Table 3 Relative Discrimination Capabilities ±2 SD (±3 SD for Sr/Zr and Fe/Cr) Numbers in parenthesis = ±3SD (±4SD for Sr/Zr and Fe/Cr) Number of Samples: 12/group (66 comparisons) No. of distinguishable pairs Ca/Nd Sn/La Mg/Cu As/Y Group 1 Distinguishable 8 (4) Elemental Profiles * Group 1 (bottle # 38 for 1 element) and Group 2 (bottles #48 and #50 for 3 elements) show some distinguishable elemental profiles from other bottles. Hf/Th Sb/Pr U/Ta Group 2 6 (0) Group 1 vs. Group 2 0 Indistinguishable (0) 1 elemental (0) profiles 1 (0) * 5 (2) * 1 (1) 1 (1) 1 (1) 1 (1) 1 (0)
Results Stage 2 (daily) Comparison Areas Group 3 7/19/05 7/30/05 Group 4 7/31/05 8/15/05 Group 5 8/9/05 8/21/05 Distinguishable Elemental Profiles (after 5 days) (after 8 days) (after 4 days) Indistinguishable elemental profiles * * 1 bottle was indistinguishable from another that was manufactured 11 days later
Results Stage 2 (daily) Table 4 Relative Discrimination Capabilities ±2 SD (±3 SD for Sr/Zr and Fe/Cr) Numbers in parenthesis = ±3SD (±4SD for Sr/Zr and Fe/Cr) Number of Samples: 12/group (66 comparisons) Comparison Distingui Indistingui No. of distinguishable pairs Elemental AreasRatios Sr/Zr Na/Al Ca/Mn shable Elementa l Profiles shable elemental profiles Group 3 30 (21) 5 (1) 3 (0) Group 4 1 (0) 1 (0) Group 5 2 (1) 14 (5) Zn/RbGroup 3 (5 15 (1) 33 (21) 50 (40) Fe/Cr7/19/05 1 (0) days*) Ba/Pb 7/30/05 22 (14) 17 (10) 34 (23) Ni/V Ti/Li Group 4 (8 Ca/Nd7/31/05 8/15/05 days) Sn/La Mg/CuGroup 5 * (4 As/Y 8/9/05 1 bottle was only distinguishable from 0 others (0) that were 8 (0) 11 days 19 (5) Hf/Th or 8/21/05 more apart. days) 1 (0) Sb/Pr 16 (6) U/Ta
Results Stage 2 (simultaneous daily) Comparison Areas Furnace 1 vs. Furnace 2 Furnace 2 vs. Furnace 3 Furnace 3 vs. Furnace 1 Furnace 1 vs. Furnace 1 Furnace 2 vs. Furnace 2 Furnace 3 vs. Furnace 3 Distinguishable Elemental Profiles Indistinguishable elemental profiles
Results Stage 2 (simultaneous daily) Table 5 Relative Discrimination Capabilities ±2 SD (±3 SD for Sr/Zr and Fe/Cr) Numbers in parenthesis = ±3SD (±4SD for Sr/Zr and Fe/Cr) Number of Samples: 6 (15 comparisons) No. of distinguishable pairs Elemental Ratios Sr/Zr Na/Al Ca/Mn Zn/Rb Fe/Cr Ba/Pb Ni/V Ti/Li Ca/Nd Sn/La Mg/Cu As/Y Hf/Th Sb/Pr U/Ta Group 6 13 (12) 15 (12) 12 (12) 13 (12) 8 (4) 8 (6) 8 (8) 2 (0) 11 (8)
Conclusion - hourly Bottles manufactured from the same furnace are typically not distinguishable when compared hourly 80 90% of bottles manufactured within a daily production lot from the same furnace share indistinguishable elemental profiles
Conclusion - daily Bottles manufactured from the same furnace between 4 and 8 days apart (11 maximum) are typically distinguishable.
Conclusion simultaneous daily Bottles manufactured at the same time from different furnaces are distinguishable. Additionally, some bottles manufactured at the same time on two consecutive days from the same furnace may be distinguishable.
Conclusion significant information Supports previous research Same case/six-pack are indistinguishable 1 Indistinguishable profiles are typically from same manufacturer 2 Gallo Glass Company Produces 800 cases/hour 1 billion bottles per year 1 Trejos, T and Almirall, J.R. Sampling strategies for the analysis of glass fragments by LA-ICP-MS Part I. Microhomogeneity study of glass and its application to the interpretation of forensic evidence. Talanta. 67(2) 388 395 (August 2005). 2 J.R. Almirall. Glass as evidence of association. In Mute Witness; When Trace Evidence Makes the Case, M. Houck (Ed.), Academic Press, San Diego, CA, USE, 139-155 (2001).
Final Conclusions Still important to collect/examine multiple fragments from question sample 1 Provides evidentiary value to forensic examinations Can distinguish between furnaces of same manufacturer Can distinguish between bottles from the same furnace after approximately 8-11 days 1 Trejos, T and Almirall, J.R. Sampling strategies for the analysis of glass fragments by LA-ICP-MS Part I. Microhomogeneity study of glass and its application to the interpretation of forensic evidence. Talanta. 67(2) 388 395 (August 2005).
Future Research Multiple manufacturer studies Bottle distribution (cases, vendors) More effective statistical studies
Acknowledgments Sacramento County District Attorney Laboratory of Forensic Services and Staff Gallo Glass Company University of California, Davis: Forensic Science Masters program California Criminalistics Institute
Contact Information Karen Harrington karenjharrington@gmail.com